건물의 전력 수요 패턴을 고려한 학습기반 예측 모델의 성능 비교 연구

초록

As electricity consumption grows every year, accurate prediction of electricity consumption is required for building energy operation management. Since the electricity consumption prediction model shows differences in predictive performance according to various factors such as building characteristics, data patterns, and prediction techniques, it is important to select an appropriate prediction model for each building to improve predictive performance. This study compares the performance of electricity consumption prediction models based on machine learning and deep learning for three buildings with different power demand scales and variability. The experiment results demonstrate that a single model doesn't show outstanding performance in all buildings and that prediction accuracy can be improved through the selection of an appropriate prediction model. The results of this study are expected to be used as guidelines for selecting an appropriate prediction model when introducing a power prediction model for a new building.

키워드